Evaluation of the Combustion Process of Pellets from Herbaceous Biomass with the Addition of Kaolin and Urea Solution in Low-Power Boilers

Abstract

In this study, an analysis was carried out of the combustion of pellets made from chamomile and English ryegrass biomass, including those with the addition of kaolin and urea, in terms of their physical and chemical properties. During combustion tests with synchronized timing, the concentrations of CO₂, CO, NO, and SO₂ in the flue gases were measured, along with the temperatures of the supplied air and the flue gases. The addition of kaolin improved combustion parameters, reduced CO emissions, and stabilized the combustion process, despite the deterioration of the mechanical durability of the pellets. Combustion in the drop-in burner (type B tests) showed higher energy efficiency (CEI) and lower flue gas toxicity (TI) than in the grate system (type A tests). The SiO₂ content in the chamomile ash explained its higher resistance to slagging, confirmed by characteristic ash temperatures. Comparison with other biofuels (straw, hay, sawdust) showed similarities or advantages in terms of reducing CO, NO, and SO₂ emissions. NO emissions were lower for pellets with urea and kaolin added, although in the case of biomass with high nitrogen content these relationships require further improvement. The research results indicate the potential of herbaceous biomass as a fuel in local heating systems. However, modification of such fuels is also associated with the need for further research on reducing emissions during unstabilized combustion phases, with particular emphasis on the ignition phase.

1. Introduction

The growing interest in renewable energy sources in recent decades has forced the intensification of research on the efficient and environmentally safe combustion of biofuels. Herbal biomass is a promising alternative to addressing the limited availability of high-quality woody biomass. However, despite the benefits of using such biomass as an energy carrier, various challenges emerge due to fluctuations in fuel properties, including moisture content, calorific value, particle size, and shape heterogeneity, as well as variations in C, H, N, S, and O content [1]. Among herbaceous biomass, a significant component is biomass from herb plantations, which cover an area of over 30,000 hectares in Poland and approximately 80,000 hectares in EU countries. Given the availability of this biomass, its use is expected to increase [2]. Due to the development of the herbal production sector, herbal production residues are available on the market. The significant availability of these raw materials creates the potential for their utilization for energy purposes. Furthermore, the use of biomass fuel produced from local raw materials also allows for the diversification of energy sources, which is a desirable approach [3]. In Poland, herbaceous biomass crops include seed grasses and chamomile. Chamomile (Matricaria chamomilla) is cultivated on an area of 1000–1200 ha in Poland, with the largest amount in the Lublin region. Under normal soil conditions, chamomile yields can reach approximately 6.0 t/ha of fresh flowers or 1.0–1.5 t/ha of dry flower heads [4,5,6]. The most popular grass in Poland is perennial ryegrass (Lolium perenne L.). This biomass is characterized by high productivity and, from an energy perspective, a high calorific value, which is important in the combustion process. However, to verify the suitability of biomass as a raw material for energy production, it is important to determine its elemental composition. The diversity of chemical properties of individual plant raw materials is most often described in the literature using elements like C, H, N, O, and Cl [1,7].

Agricultural biomass, especially herbaceous biomass, has a chemical composition that may undermine its suitability for low-emission heating systems. High nitrogen, sulfur, and chlorine content not only increases the risk of secondary pollutant emissions (NOx, SOx, and particulate matter) but also leads to operational problems, such as the corrosion of combustion systems. This suggests that sustainable biomass use cannot be limited to CO₂ balance alone—it must consider the full emission profile and the impact on infrastructure [8,9].

Paradoxically, co-firing biomass with fossil fuels, such as low-nitrogen coke, can improve emission parameters, but it also pushes away from the idea of full decarbonization. In this view, biomass ceases to be a zero-emission solution and becomes part of a technological compromise—a transitional measure, not the ultimate solution [10].

Further development of this technology should therefore focus not only on optimizing combustion devices but also selecting raw materials and modifying biomass processing processes to minimize secondary emissions and infrastructure risks. Sustainable bioenergy must be defined not by the appearance of carbon neutrality but by the full life cycle of the raw material and its impacts [11].

Greater efficiency in biomass combustion, accompanied by reduced gaseous and particulate emissions, can be achieved through precise staging of combustion air supply. Air staging is a well-established method for reducing NOx emissions. When combined with accurate control of residence time and the primary air excess ratio, it can deliver substantial NOx reduction, often in the range of 50–80%. The volatile content matter of the fuel also plays a significant role in determining the extent of NOx reduction [12].

However, the processing and use of herbaceous biomass is problematic. Herbaceous biomass varieties are typically characterized by higher silicon and alkali metal content, which is considered the main factor contributing to low ash melting temperatures, resulting in an increased risk of slagging and deposits on the convection surfaces of the heating device [13].

Although biomass as a fuel has a relatively small impact on the CO₂ emission balance, its application in heating devices faces several technological challenges. As highlighted in previous studies [14], these challenges can be addressed by enhancing the chemical and physical properties of solid biofuels and by introducing additives that improve the combustion process—particularly in low-power heating systems—while simultaneously reducing the emission of gaseous products into the atmosphere.

One of the key problems is the deposition of ash on heat exchange surfaces and the formation of slags and sinters, which leads to reduced equipment efficiency, increased operating costs, and potential damage to the installation [15,16]. Ash generated during biomass combustion has significantly different physicochemical properties than that generated during the combustion of fossil fuels. First of all, it contains a higher share of alkalis (K, Na), which results in a lower melting point and an increased tendency to form slag deposits and glassy deposits [17,18]. Additionally, thermal processes occurring during biomass combustion lead to the sublimation and condensation of compounds, such as KCl and K₂SO₄, which actively participate in high-temperature corrosion [19,20,21].

In light of the problems above, intensive research is being conducted on the use of mineral additives, such as kaolin, and chemical reagents, such as urea, in order to modify ash’s properties and improve combustion parameters. Kaolin (Al₂Si₂O₅(OH)₄), a natural aluminosilicate, has the ability to react with volatile compounds of potassium, sodium, and chlorine, leading to the formation of stable, high-melting aluminosilicates (e.g., leucite, nepheline), which significantly reduce slagging and ash deposition [22].

Moreover, the latest literature reports also indicate the participation of kaolin in NOₓ emission reduction processes. This is due to its catalytic and adsorption effect; aluminosilicates can sorb precursor forms of nitrogen at high temperatures and also mediate the conversion of NO to N₂, especially in the presence of water vapor and with an appropriate thermal regime. Such properties are primarily attributed to the surface reactivity of the Al₂O₃ phase and the microporous mineral structure [23,24,25]. Some studies suggest that the presence of kaolin can contribute to the reduction of NO_x emissions by supporting the reaction of nitrogen oxides’ reduction. Furthermore, the addition of kaolin to fuel can change combustion characteristics, including flame temperature. Reducing the temperature can reduce the formation of NO_x, which is mainly formed at high temperatures. Due to its porous structure, kaolin can adsorb some pollutants, including CO and NO_x, which potentially reduces their emission to the atmosphere. Some studies have analyzed the effect of various mineral additives, including kaolin, on pollutant emissions during the combustion of solid fuels. The results suggest that the presence of kaolin can reduce NO_x emissions, but these effects are dependent on many factors, such as fuel composition, combustion conditions, and additive proportions [26,27,28,29]. Kaolin is sometimes used as a carrier in catalysts for reducing exhaust emissions. Its structure and chemical properties can support NO_x reduction processes in the presence of appropriate active catalytic components. It is important here that the porous structure of kaolin allows for good dispersion of active catalytic components, and the presence of Brønsted acid sites promotes ammonia adsorption, which is important in the SCR process. Brønsted acid sites can donate a proton (H⁺). In the context of mineral materials, such as aluminosilicates, these are most often –OH groups bound to aluminum or silicon atoms. In the reaction environment, they can protonate reactive molecules, such as ammonia (NH₃), facilitating its adsorption, and participate in the activation of reagents, e.g., by protonating NO or its derivatives [30,31,32]. Additionally, the nitrogen content of such herbaceous biomass is problematic, resulting in increased NO_x emissions during combustion [33]. The use of NO_x-reducing chemicals is helpful here. Urea [34] is one such compound. It is commonly applied as a NOₓ-reducing agent in selective non-catalytic reduction (SNCR) processes, and it can also influence the chemical composition of exhaust gases by limiting the formation of alkali metal chlorides and sulfates while providing anticorrosive protection for heating equipment [35,36,37]. It is an organic chemical compound constituting a diamide of carbonic acid. Introducing an appropriate additive to the pelleting and direct fuel combustion processes reduces the amount of dioxins produced. Due to its activity in the DeNO_x process, a dioxin inhibitor can reduce the emission of nitrogen oxides [34]. Synergistic use of kaolin and urea can therefore not only improve the efficiency of the combustion process but also significantly reduce the emissions of nitrogen oxides, chlorine, and particulate matter.

The document regarding the monitoring of gas emissions (CO₂, NO_x) from technological processes, exhaust gas treatment, and the use of additives is Regulation (EU) 2023/956 of the European Parliament and of the Council [38]. This document covers all stationary fuel combustion units, including boilers, burners, furnaces, incinerators, kilns, etc. Although there are indications that kaolin may affect CO and NO_x emissions, there is no clear evidence to confirm this effect under operating conditions for low-power heating appliances. Therefore, research is needed to precisely determine the mechanisms of action of kaolin in combustion processes and its potential benefits in reducing pollutant emissions.

Hence, the aim of the research is to assess energy and ecological efficiency in terms of changes in the combustion process and emissions of exhaust gases, such as CO, NO, and SO₂, in a heating device with a different combustion chamber design in which biofuels in the form of pellets made from herbaceous biomass, with the addition of kaolin or urea, were used. This article focuses on the analysis of the effect of such additives on the course of the biofuel combustion process and the related characteristic features of ash temperatures.

2. Materials and Methods

2.1. Fuel

For the study, 8 types of plant pellets were selected, which were made from plant raw materials in the form of waste biomass after cleaning English ryegrass seeds and chamomile inflorescences, which consisted of small elements of stems, seeds, and mineral impurities, as well as an admixture aimed at preventing ash caking and reducing NO_x using powdered porcelain clay with a relative density of 2.6 g·cm⁻³, pH 4.5, cosmetic purity >95%, and 32.5% ± 0.7% aqueous urea solution. Four fractions were isolated from the prepared raw materials within a given species. The first was 100% raw material for chamomile waste 100R and for ryegrass designated as 100T. The second comprised mixtures with 95% plant raw material and 5% kaolin, designated as 95R5G and 95T5G, respectively. The third one consisted of mixtures of 95% raw material and 5% water solution with 32.5% urea content, designated as 95R5M and 95T5M. The fourth one consisted of mixtures of 90% raw material and 5% kaolin and urea solution, designated as 90R5G5M and 90T5G5M. The agglomeration of raw materials and their mixtures was carried out using a granulator equipped with a stationary flat die and rotating pressure rollers, driven by an electric motor with a rated power of 7.5 kW. The granulator die had 8 mm diameter holes, and the pellets produced had a length-to-diameter (L:D) ratio of 3.125. An evaluation of the potential for pelleting selected herbal biomass with the addition of kaolin and a urea solution is presented in [39].

Research Methods to Analyze the Fuels Used

Before the combustion tests, the produced pellets were analyzed for the following content:

-

Moisture with an accuracy of 0.01%, using the dryer-weighing method according to the PN-EN ISO 18134-3:2015 standard [40].

-

Volatile matter (VM), by weight according to (PN-G-04516:1998) [41].

-

Carbon (C), hydrogen (H), nitrogen (N)—automatic analysis according to (CEN/TS 15104:2006) for three repetitions from a sample taken from crushed pellets [42].

-

Sulfur (S)—in accordance with the requirements of PN-G-04584:2001, for three repetitions from a sample taken from crushed pellets [43].

-

Heat of combustion (HHV) using a Parr 6400 isoperibolic calorimeter (Parr Instruments, Moline, IL, USA), according to PN-EN ISO 18125:2017 [44].

-

Ash (AC) using a laboratory furnace according to PN-EN ISO 18122:2016 [45].

In addition, in accordance with the specificity of the research, additional analyses and determinations were performed for the collected research material:

-

Chemical composition of ash, which was determined through plasma spectrometry using the Thermo iCAP 6500 Duo ICP device (Thermo Fisher Scientific Inc., Waltham, MA, USA), including chloride content, determined using the titration method according to PN-EN 196-2:2006 [46].

-

Characteristic ash melting temperatures determined according to the requirements of the PN-EN ISO 21404:2020-08 method [47].

The chemical characterization of the prepared research material was supplemented by microscopic analysis using a Keyence VHX-X1 digital microscope (Keyence International, Mechelen, Belgium), which verified the presence of the additives used on the surface and between the fibers of the densified biomass based on an image on a 50 μm scale.

2.2. Research Stand

As in the work of Dula et al. [16], the combustion tests were divided into two parts: type A tests and type B tests. The study was conducted in two phases, each utilizing a different combustion system. In the initial phase (type A tests), combustion trials were carried out using a conventional grate boiler rated at 10 kW thermal output and 80% efficiency. The boiler featured a cuboid combustion chamber (0.26 × 0.30 × 0.45 m) enclosed by a water jacket. Primary air was supplied beneath the grate at a velocity of 1 m·s⁻¹ via a regulated fan. The exhaust gases exited through a 3 m tall chimney with a 130 mm internal diameter.

For the second phase (type B tests), the setup was upgraded with an automatic drop-in burner mounted in the boiler door. This burner system used a screw feeder to deliver pelletized biomass fuel from a hopper, directing it into the combustion chamber via a chute. Inside of the burner, the screw conveyed fuel to the combustion zone, with a ceramic deflector plate positioned above to optimize flame geometry and heat retention. An integrated electric igniter enabled automated ignition, while fuel feed rate, air supply, and operation cycles were controlled electronically.

In both setups, the test stand was outfitted with instrumentation for thermal performance assessment, including a circulation pump, flow meter, pressure sensor, and temperature probes measuring water at the inlet and outlet of the boiler loop.

A schematic diagram of the complete test system, illustrating both combustion configurations, is presented in Figure 1.

2.3. Course of Combustion Tests

The experimental procedure consisted of two main phases. In the first phase, the combustion characteristics of fuel pellets were examined under standardized batch conditions. Each test used 1 kg of pellets, ignited by placing them on an established ember bed. Combustion continued until the exhaust gas temperature dropped to 200 °C, indicating the end of the reaction. This setup reflected typical user operation in residential heating systems. Flue gases were sampled directly from the chimney (Figure 1).

The second phase focused on the three-stage operation of an automated pellet burner. The initial stage involved automatic ignition, during which the fan first cleared ash from the grate, followed by pellet dosing and activation of the electric igniter and fan. The fuel was heated until flame detection by an optical sensor. Throughout this stage, the fan operated continuously, while pellets were intermittently supplied by a screw feeder from the hopper. Upon flame detection, the system transitioned to nominal operation (stage II), during which the boiler heated the water jacket to the target temperature. Testing in this phase involved burning a 1 kg portion of the selected fuel, with a dosing cycle of 8 s on and 10 s off and a fan power set to 30%. After combustion, the system was cooled and cleaned before the next cycle.

In both setups, combustion duration was recorded, while exhaust gas composition and temperature were continuously monitored. The measuring probe was connected to an exhaust gas dryer, which directed the gases to an exhaust gas analyzer. A portable analyzer equipped with non-dispersive infrared (NDIR) sensors was used to measure CO, CO₂, NO, and SO₂ concentrations. Temperature measurements were taken using a K-type thermocouple integrated into the exhaust gas sampling probe. The analyzer’s detailed technical specifications are provided in

Table 1. Technical specifications of the gas analyzer equipment.

Gas	Measurement Range	Accuracy	Resolution	Type of Measurement
CO2	0–25%	±3%	0.01%	NDIR
CO	0–20,000 ppm	±3 ppm	1 ppm	NDIR
NO	0–5000 ppm	±3 ppm	1 ppm	NDIR
SO2	0–5000 ppm	±3 ppm	1 ppm	NDIR
Tgas	−10 ÷ 1000 °C	±2 °C	0.1 °C	Type K thermocouple

.

The test results recorded by the analyzer were transferred to a PC, where their visual distribution was examined using Microsoft Excel. Combustion phases were identified based on changes in CO and CO₂ emissions, following the methodology described by Dula et al. [3]:

Phase I—ignition: Flameless release of volatile fuel components, reaching a maximum CO concentration.

Phase II—main combustion: Visible flame formation and rapid afterburning of volatile compounds, leading to a minimum CO level in the exhaust gases.

Phase III—afterburning: Disappearance of the flame, oxidation of remaining fuel in the ember bed, and a renewed increase in CO emissions.

For each test variant, three repeated measurements were carried out, and the arithmetic mean was calculated for the identified phases.

The measured CO, NO, and SO₂ concentrations in the flue gases were normalized to the dry flue gas volume with 10% oxygen content under standard conditions (0 °C, 1013 mbar), expressed in mg·m⁻³, in accordance with the PN-EN 303-5:2002 standard [48]. The combustion efficiency index (CEI) was calculated using the formula [16]

CEI = 100 − q_A (%)

(1)

where

• q_A—chimney loss calculated according to Formula (2).

$${\mathrm{q}}_{\mathrm{A}}=({\mathrm{T}}_{\mathrm{gas}}-{\mathrm{T}}_{\mathrm{amb}}) \times ({\displaystyle \frac{A_{\mathit{1}}}{{CO}_{\mathit{2}}} + \mathrm{B}) (\%)}$$

(2)

where

• T_gas—flue gas temperature (°C);
• T_amb—air temperature at the boiler inlet (ambient temperature) (°C);
• CO₂—carbon dioxide concentration in flue gas (%);
• A₁, B—Siegert coefficients characteristic of biomass, A₁ = 0.65, B = 0.

The toxicity index (TI) was calculated based on Formula (3) as the quotient of carbon monoxide (CO) concentration in exhaust gases to carbon dioxide (CO₂) concentration in exhaust gases [16]:

$$\mathrm{TI} = {\displaystyle \frac{CO}{{CO}_2}} (\text{-})$$

(3)

where

• CO—concentration of carbon monoxide in exhaust gases (%);
• CO₂—concentration of carbon dioxide in exhaust gases (%).

2.4. Statistical Analysis

To assess the distribution of the obtained data and ensure the validity of further parametric analyses, the Shapiro–Wilk test was applied due to its high power for detecting deviations from normality in small and moderate sample sizes. To evaluate the assumption of homogeneity of variances required for ANOVA, the Brown–Forsythe test was used, which is more robust to departures from normality compared to Levene’s test.

Given the experimental nature, both one-way ANOVA (for analyzing the physical and chemical composition of biofuels or ashes) and factorial ANOVA (to evaluate the interaction between fuel type and combustion system and combustion products) were employed. These tests were appropriate for comparing group means when normality and variance homogeneity conditions were met.

Tukey’s HSD post hoc test was used to identify statistically homogeneous groups, as it controls the familywise error rate and is suitable for all pairwise comparisons following ANOVA with equal group sizes or balanced designs.

To assess linear relationships between continuous variables (e.g., combustion parameters, emissions), Pearson’s correlation coefficient was applied under the assumption of approximate normality and linearity of relationships.

For multivariate pattern recognition and grouping of cases with similar profiles, hierarchical cluster analysis was performed. Variables were standardized to avoid bias due to scale differences. Ward’s linkage method was selected as it minimizes total within-cluster variance, and Euclidean distance was used as a dissimilarity metric appropriate for continuous, standardized data. The observed differences were considered statistically significant at a significance level of p < 0.05.

3. Results

3.1. Results from the Combustion Tests

As shown in the work of Dula et al. [39], the diameter of the pellets used in the study was 8 mm, and, as a variable, it did not show any variability. On the other hand, the length of the pellets was primarily influenced by the type, moisture content, and bulk density of the raw materials. The shortest pellets were obtained from raw materials without additives, and the longest were obtained in a mixture of chamomile biomass with porcelain clay (kaolin) and urea solution. The differentiation of the dimensions of the produced granules also resulted in a change in bulk and volumetric density. The results of the bulk density tests of the pellets produced for the adopted variants, similarly to the previously analyzed parameters, indicated that the lowest bulk density was characteristic of pellets produced with the addition of kaolin (from 763 to 875 kg·m⁻³), while pellets without additives and with urea solution were characterized by a slightly greater density (from 1022 to 1120 kg·m⁻³). At the same time, for the analyzed agglomerates, the increase in the amount of additives used caused a decrease in the value of this parameter from about 470 to 380 kg·m⁻³. On the other hand, the energy characteristics and elemental composition of these agglomerates are presented in

Table 2. Energetic properties of the pellets’ combustion.

Type of Pellets	MC (%)	HHV (MJ·kg−1)	AC (%)	VM (%)
100R	9.47 ± 0.05 a	18.21 ± 0.08	10.38 ± 0.49 a	66 ± 2 a
100T	8.43 ± 0.07	17.49 ± 0.10 a	11.01 ± 0.49 a.b	71 ± 3 a.b
95R5M	9.98 ± 0.06 b	18.03 ± 0.02	10.78 ± 0.11 a.b.c	68 ± 2 a.b.c
95T5M	9.26 ± 0.03 a.c	17.46 ± 0.01 a.b	11.24 ± 0.34 b.c	73 ± 3 b.c.d
95R5G	9.95 ± 0.33 b	17.68 ± 0.09 c	13.10 ± 0.09 d	63 ± 2 a.c.e
95T5G	9.43 ± 0.20 a.c	16.90 ± 0.05d	14.82 ± 0.05 e	67 ± 3 a.b.c.d.f
90R5G5M	10.69 ± 0.10 d	17.57 ± 0.02 a.b.c	13.46 ± 0.03 d.f	60 ± 2 a.e.g
90T5G5M	10.72 ± 0.03 d	16.77 ± 0.03 d	14.16 ± 0.15 e.f	64 ± 3 a.c.e.f.g

Note: R—chamomile biomass; T—ryegrass biomass; G—porcelain clay (kaolin); M—urea solution; MC—moisture content; HHV–combustion heat; AC—ash content; VM—volatile matter; ^a–g—homogeneous group, ±standard deviation.

and

Table 3. Elemental composition of the pellets’ combustion.

Type of Pellets	C (%)	H (%)	N (%)	S (%)
100R	40.18 ± 0.40	5.06 ± 0.07 a	2.82 ± 0.15 a	0.01 ± 0.01
100T	36.77 ± 0.56 a	5.40 ± 0.16 b	2.90 ± 0.08 a.b	0.04 ± 0.02
95R5M	38.50 ± 0.38b	4.91 ± 0.06 a.c	3.43 ± 0.14 c	0.01 ± 0.01
95T5M	35.25 ± 0.53 c	5.24 ± 0.15 a.b.c.d	3.50 ± 0.07 c.d	0.04 ± 0.02
95R5G	38.17 ± 0.38 b	4.80 ± 0.06 a.c.e	2.68 ± 0.14 a.b.e	0.01 ± 0.01
95T5G	34.93 ± 0.53 c	5.13 ± 0.15 a-f	2.75 ± 0.07 a.b.e	0.04 ± 0.02
90R5G5M	36.48 ± 0.36 a.c	4.66 ± 0.06 c.e.g	3.15 ± 0.13 c.d.f	0.01 ± 0.01
90T5G5M	33.42 ± 0.50	4.95 ± 0.14 a.c.d.e.f.g	3.36 ± 0.07 c.d.f	0.03 ± 0.01

Note: R—chamomile biomass; T—ryegrass biomass; G—porcelain clay (kaolin); M—urea solution; C—carbon content; H—hydrogen content; N—nitrogen content; S—sulfur content; ^a–g—homogeneous group, ±standard deviation.

. From the perspective of the energy assessment of the prepared fuel, the addition of kaolin and/or urea solution caused a decrease in the combustion heat of pellets. Additionally, kaolin as a mineral additive increased the already high ash content. The relatively high value of this parameter in the analyzed biomass/pellets is due to the fact that it is a post-production raw material—a by-product—contaminated with a mineral fraction, such as soil. The presence of additives in the form of kaolin or urea is visible in microscopic images on a scale of 50 μm, which show mineral inclusions (kaolin) and a network with urea crystals (Figure 2).

Table 4. Chemical properties of ashes’ biofuel combustion in %.

Symbol	100R	100T	95R5M	95T5M	95R5G	95T5G	90R5G5M	90R5G5M
SiO2	60.63 ±1.9 a	43.2 ±1.4 b	57.59 ±1.8 a,c	41.04 ±1.3 b,d	58.69 ±1.8 a,c,e	42.14 ±1.3 b,d	55.66 ±1.7 c,e,f	39.98 ±1.3 b,d,f
P2O5	6.43 ±0.4 a	8.42 ±0.5 b	6.11 ±0.38 a	7.99 ±0.47 b,d	6.10 ±0.38 a,c,e	7.99 ±0.47 b,d,e	5.78 ±0.36 a,c,d	7.57 ±0.45 a,b,d,e
K2O	17.15 ±0.4 a	21.5 ±0.4 b	16.29 ±0.38 a,c	20.42 ±0.38 b,d	16.29 ±0.38 a,c,e	20.42 ±0.38 b,d,f	15.43 ±0.36 c,e	19.3 5 ± 0.36 d,f
CaO	2.54 ±0.2 a	11.2 ±0.3 b	2.41 ±0.19 a,c	10.64 ±0.28 b,d	2.41 ±0.19 a,c,d,e	10.64 ±0.28 b,f	2.28 ±0.18 a,c,e	10.08 ±0.27 d,f
MgO	0.35 ±0.02 a	4.81 ±0.1	0.33 ±0.02 a,b	4.57 ±0.09 c	0.33 ±0.02 a,b,d	4.56 ±0.09 c	0.31 ±0.02 a,b,d	4.32 ±0.09
Na2O	1.23 ±0.1 a	0.51 ±0.05 b	1.17 ±0.09 a,c	0.48 ±0.05 b,d	1.16 ±0.09 a,c,e	0.48 ±0.05 b,d,f	1.10 ±0.09 a,c,e	0.45 ±0.05 b,d,f
SO3	2.62 ±0.3	2.8 ±0.3	2.49 ±0.28	2.66 ±0.28	2.48 ±0.28	2.66 ±0.28	2.35 ±0.27	2.52 ±0.27
Fe2O3	0.96 ±0.1	0.76 ±0.1	0.91 ±0.09	0.72 ±0.09	0.91 ±0.09	0.72 ±0.09	0.86 ±0.09	0.68 ±0.09
Al2O3	3.44 ±0.2 a	1.21 ±0.1 b	3.27 ±0.19 a	1.15 ±0.09 b	4.31 ±0.19 c	2.19 ±0.09 d	4.14 ±0.18 c	2.13 ±0.09 d
Mn3O4	0.53 ±0.05 a	0.26 ±0.03 b	0.50 ±0.05 a,c	0.25 ±0.03 b,d	0.50 ±0.05 a,c,e	0.24 ±0.03 b,d,f	0.47 ±0.05 a,c,e	0.23 ±0.03 b,d,f
BaO	0.07 ±0.01	0.05 ±0.01	0.07 ±0.01	0.047 ±0.09	0.06 ±0.01	0.04 ±0.01	0.06 ±0.01	0.04 ±0.01
TiO2	0.23 ±0.02 a	0.12 ±0.01 b	0.22 ±0.02 a,c	0.11 ±0.09 b,d	0.21 ±0.02 a,c,e	0.11 ±0.01 b,d,f	0.20 ±0.02 a,c,e	0.11 ±0.01 b,d,f
SrO	0.04 ±0.01	0.06 ±0.01	0.04 ±0.01	0.06 ±0.09	0.03 ±0.01	0.05 ±0.01	0.03 ±0.01	0.05 ±0.01
Cl	0.17 ±0.03 a	1.29 ±0.05 b	0.16 ±0.03 a,c	1.23 ±0.05 b,d	0.16 ±0.02 a,c,e	1.22 ±0.05 b,d,f	0.15 ±0.03 a,c,e	1.16 ±0.05 d,f
CO2	3.03 ±0.1	3.04 ±0.1	2.88 ±0.09	2.89 ±0.09	2.87 ±0.09	2.88 ±0.09	2.72 ±0.09	2.73 ±0.09

Note: ^a–f homogeneous group, ±standard deviation.

present the chemical characteristics of the ash used in the pellet tests. Upon comparing the chemical composition of the ash of both raw materials used in the study, a high SiO₂ content in chamomile (100R) is clearly visible, 60.63% vs. 43.2% (100T), which suggests significant amounts of silicon, which may indicate not only the accumulation of this compound in the biomass but also the contamination of such biomass (e.g., with soil). Additionally, both raw materials have very high contents of potassium (K₂O) and phosphorus (P₂O₅), amounting to 17.15% and 21.5% for K₂O and 6.43% and 8.42% for P₂O₅, respectively. Nevertheless, taking into account their melting temperatures (740 and 360 °C), the higher shares of these compounds in ryegrass ash indicate its greater susceptibility to ash melting [49]. In addition, the ash from ryegrass biomass, compared to chamomile, contains higher contents of CaO, MgO, Fe₂O₃, and Al₂O₃, which suggests that this plant accumulates alkaline earth elements and metals better. The addition of urea solution to the biomass under consideration, to a small extent resulting from the percentage shares of the mixtures, reduced the amount of individual mineral compounds in the ash of the pellets produced with an admixture of this compound. On the other hand, the addition of kaolin, due to its content of aluminosilicates, slightly increased the proportion of Al₂O₃ and SiO₂.

The characteristic melting temperatures of ash from chamomile and ryegrass biomass, taking into account the admixtures used in the tests, are presented in

Table 5. Characteristic melting temperatures of ash of the pellets’ combustion.

Type of Pellets	Sintering Temperature SST (°C)	Deformation Temperature DT (°C)	Melting Temperature HT (°C)	Flow Temperature FT (°C)
100R	1180 ± 20 a	1260 ± 50 a	1280 ± 20 a	1290 ± 10 a
100T	1010 ± 20 b	1090 ± 50 b	1130 ± 20 b	1180 ± 10 b
95R5M	1160 ± 20 a,c	1250 ± 50 a	1270 ± 30 a	1280 ± 10 a,c
95T5M	1000 ± 20 b	1080 ± 50 b	1120 ± 20 b	1170 ± 10 b
95R5G	1200 ± 30 a,c,d	1440 ± 50 c	1460 ± 20 c	>1500 ± 10
95T5G	1380 ± 20 e	1430 ± 50 d	1440 ± 20 c,d	1460 ± 10 d
90R5G5M	1200 ± 30 a,c,d	1430 ± 50 c,d,e	1460 ± 30 c,d,e	>1500 ± 20 c,e
90T5G5M	1400 ± 20 e	1430 ± 50 c,d,e	1440 ± 20 c,d,e	1470 ± 20 c,d,e

Note: ^a–e homogeneous group, ±standard deviation.

. Ryegrass ash (100T) has significantly worse melting properties than chamomile ash (100R). The FT flow temperature for ryegrass is only 1180 °C, and the DT deformation temperature is already at 1090 °C, hence the high risk of slagging at temperatures typical of grate or fluidized bed boilers. At the same time, chamomile has temperatures ~100–200 °C higher, which is why this ash is more stable. Ryegrass is more problematic, probably due to the lower content of SiO₂ in the stabilizing form and the presence of reactive K, Mg, and P. The differences between 100R or 100T and 95R5M or 95T5M—10–20 °C in melting temperatures—are not significant. If the ash temperature changes after adding urea, it is instead an indirect effect, e.g., a change in ash formation through other chemical reactions.

Porcelain clay (kaolin-G) significantly improves ash properties in the case of chamomile by +180 °C (DT), +180 °C (HT), and +>210 °C (FT). In turn, for ryegrass ash, these increases are even more spectacular, e.g., FT increases from 1180 °C to 1460 °C. It is probable that kaolin forms more stable phases with K and P, like potassium aluminosilicates. The combination of clay and urea (G + M) gives a synergistic effect. For both types of biomass, the fusion temperature (FT) exceeds 1450 °C, which is well above the safe combustion threshold (FT > 1250 °C), even for steel grates [37].

The main objective of this study, considering the diverse physical and chemical properties of green biomass pellets made from chamomile and English ryegrass, was to compare the combustion process in two types of heating devices: test A, a unit where combustion occurred periodically on a grate with air supplied from below, and test B, a unit with automatic feeding to a drop-in burner. The analysis focused solely on the second phase of combustion, excluding the ignition phase (I) and the final fuel burnout phase (III).

For grate combustion, the second phase lasted from 308 s (90R5G5M) to 580 s (95R5G). In the drop-in burner, the corresponding phase lasted from 212 s (95R5G) to 432 s (100R and 100T). Figure 3 presents the CO₂ concentration in the exhaust gases and the exhaust gas temperature.

During the type A combustion tests, the CO₂ content when supplying the boiler by periodically placing fuel on the grate for pellets generally ranged between 7 and 9%. In contrast, under these test conditions, the combustion of pellets from English ryegrass biomass modified with additives (95T5M, 95T5G, 90T5G5M) was characterized by significantly lower CO₂ emissions of about 5%. The change in the feed system to the boiler during the type B tests resulted in increased combustion intensity while smoothing its course. The CO₂ content in these conditions was in a narrow range, from less than 11 to less than 14% (Figure 3).

The average exhaust gas temperature varied greatly. The lowest values, from 220 to 280 °C, were recorded for tests A and B, in which pellets from English ryegrass biomass modified with kaolin and urea additives were burned. On the other hand, the exhaust gas temperature in the range of 300–380 °C was characteristic of pellets from chamomile biomass, regardless of the variant of production and combustion.

The recorded emissions of exhaust gas components, such as CO, NO, and SO₂, are presented in Figure 4.

During combustion tests A and B, carbon monoxide emission was the highest for pellets without kaolin (100R, 100T, 95R5M, 95T5M) and ranged from 8 to 12 g·m⁻³ at 10% O₂. The emission during the combustion of pellets from English ryegrass was close to the lower limit of this range. The addition of kaolin caused CO emission during tests A and B to be lower than 8 g·m⁻³ at 10% O₂. It is important to note here that during type B tests, regardless of the raw material but provided that kaolin was present, it was usually around 2 g·m⁻³ at 10% O₂. These results indicate the unevenness and randomness of combustion of the considered solid biofuels on the grate. Automatic fuel feeding and its combustion in smaller portions allow for the reduction of this gas emission, especially in mixtures with the addition of kaolin (Figure 4).

During tests A and B, regardless of the combustion system, NO emission differences were observed. The emission of this gas was in the range of 150–450 mg m⁻³ at 10% O₂. At the same time, it was observed that when burning the pellets produced with additives in the form of urea solution or kaolin, NO emission was in the range of 150–250 mg m⁻³ at 10% O₂ (Figure 4). During the tests, significant variations in SO₂ emissions were also observed. The highest values, 4–5 times higher than the average of 200 mg·m⁻³ at 10% O₂, were observed for chamomile biomass pellets during type B tests (Figure 4).

The combustion efficiency and toxicity index are shown in Figure 5.

3.2. Results’ Statistical Analysis

The obtained results of the one-factor ANOVA analysis of variance for technical characteristics, the chemical composition of pellets, and the characteristic temperatures and chemical composition of ash are presented in

Table 6. Single factor analysis of variance (ANOVA) for the energetics and chemical fuel properties.

Source of Variation	MC, %	HHV, MJ·kg−1	AC, %	VM, %	C, %	H, %	N, %	S, %
Raw material (A)	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *	0.026 *
Source of variation	SiO2, P2O5, K2O, CaO, MgO, Na2O Al2O3, Mn3O4, TiO2, Cl			SO3	Fe2O3	BaO	SrO	CO2
Raw material (A)	<0.001 *			0.675 *	0.013 *	0.016 *	0.017 *	0.007 *
Source of variation	Sintering temperature SST (°C)		Deformation temperature DT (°C)		Melting temperature HT (°C)		Flow temperature FT (°C)
Raw material (A)	<0.001 *		<0.001 *		<0.001 *		<0.001 *

* significant values (p < 0.05).

, while the factor analysis for parameters and combustion products is presented in

Table 7. Factor analysis of variance (ANOVA) for the combustion products.

Source of Variation	CO2	Tgas	CO	NO	SO2	CEI	TI
Combustion system (A)	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *
Raw material (B)	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *
A × B	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *	<0.001 *

* significant values (p < 0.05).

.

Statistical analysis in the group of fuel energy characteristics showed, as a rule, significant differences between the mean values of all analyzed biofuel parameters, and only for sulphur content was such a difference not observed, with probability p = 0.026 (Table 6). On the other hand, among the analyzed characteristics of ash from pellets used in the study, the chemical composition did not differ significantly for the content of SO₃, Fe₂O₃, BaO, SrO, or CO, with probabilities of 0.675, 0.013, 0.016, 0.017, and 0.007, respectively. The remaining ash components showed significant differentiation (p < 0.001) between the analyzed pellets (Table 6). Additionally, during these statistical analyses, significant differences (p < 0.001) were also obtained for characteristic ash temperatures (Table 6).

Statistical analysis also showed significant differences between the mean values of all analyzed exhaust gas parameters (p < 0.001). However, Tukey’s post hoc tests indicated no statistically significant differences between emissions during combustion of the following pairs: 100R test A and 100R test B; 100T test A and 95T5M test A and 100T test B and 95R5M test B and 95T5M test B; and 95T5M test A. The cluster analysis for the tested pellet types in the considered groups of features is presented in Figure 6, Figure 7 and Figure 8. In the group of energetic and chemical features of raw materials and ash, there were noticeable groups differentiated within the raw materials, thus separating pellets with the addition of kaolin or without this compound (Figure 6 and Figure 7, respectively). However, in the combustion tests, the isolated clusters were not as regular, although the bond distance for the two groups was very similar (Figure 8).

Pearson correlation analyses for individual types of combustion tests are presented in

Table 8. The pellets periodically on the grate.

Variable	C	H	N	S	AC	VM
Type A tests
CO2	0.647701 *	−0.030709	−0.352532 *	−0.398294 *	−0.474339 *	0.014296
Tgas	0.693650 *	−0.075218 *	−0.340931 *	−0.470904 *	−0.500510 *	−0.056949
CO	0.210733 *	0.393075 *	0.323490 *	−0.015180	−0.526565 *	0.180340 *
SO2	−0.013679	0.163047 *	0.147837 *	−0.107534 *	−0.039472	−0.106421 *
NO	−0.249189 *	0.127415 *	0.484259 *	0.131485 *	−0.151346 *	0.036686
ETA	0.207041 *	0.389572 *	0.330897 *	−0.015614	−0.085083 *	0.114802 *
TI	0.154378 *	0.062261	−0.175651 *	−0.020596	−0.525490 *	0.175978 *
Type B tests
CO2	0.358159 *	−0.148986 *	−0.391896 *	−0.468039 *	0.099790 *	−0.305301 *
Tgas	0.554977 *	−0.113706 *	−0.250692 *	−0.705135 *	−0.091037 *	−0.466488 *
CO	0.454569 *	0.513440 *	−0.066281	0.035572	−0.674822 *	0.432395 *
SO2	0.603930 *	−0.127452 *	−0.364059 *	−0.749998 *	−0.099833 *	−0.515665 *
NO	0.191253 *	0.179224 *	−0.226699 *	−0.160572 *	−0.069974	0.003889
ETA	0.456325 *	0.479496 *	−0.108679 *	0.055930	0.157455 *	0.343680 *
TI	−0.403514 *	0.045855	0.033093	0.508712 *	−0.678190 *	0.471703 *

* marked correlation coefficients are significant, with p < 0.05.

. The analysis of test results performed using Pearson correlations showed clear differences in the relationships between fuel composition and process parameters and emissions depending on the combustion technology. In type A tests (grate combustion), the relationships are generally weaker and less clear, which may result from greater variability of combustion conditions (e.g., uneven fuel distribution, local underburning). For example, the correlation between carbon content in the fuel and CO₂ emissions is clear in both cases but stronger in the grate system, which may indicate that for this technology, carbon content in the fuel translates more directly into CO₂ emissions.

In turn, in type B tests (drop-in burner), correlations are stronger and more consistent—for example, CO or SO₂ emissions or exhaust gas temperature show strong links with the C, H, and S content in the fuel. This suggests that combustion conditions in burner systems are more controlled and uniform, which is reflected in clearer relationships between fuel composition and combustion effects. On the other hand, NO emission has very low correlations with most variables, which may mean that its emission depends mainly on other, more complex factors and not the nitrogen content in the fuel. The inverse correlation between ash content (AC) and gas emissions in type B tests is also interesting; this may indicate that a higher content of non-flammable components suppresses the combustion process, reducing the intensity of the reaction. Hence, this part of the analysis of results emphasizes that the type of combustion technology significantly affects the relationships in which the fuel’s composition translates into emissions and boiler operating parameters. Systems using automatic fuel supply to the combustion chamber show stronger coupling, which may suggest that they are more susceptible to optimization based on fuel composition.

4. Discussion

The biofuels used in the study in the form of pellets belong to the group of biofuels made from herbaceous biomass due to the characteristics of the raw materials used. In terms of chemical characteristics, as described in the work of Dula et al. [39] for [50,51,52], the collected raw materials were similar to the best types of biomass intended for energy purposes. It is also important that the pellets produced with porcelain clay did not meet the provisions of the relevant quality standards PN-EN ISO 17225-6:2014-08 [53] in terms of mechanical durability. However, the conducted studies clearly indicate that the chemical and physical properties of the biomass of chamomile (100R) and English ryegrass (100T), as well as the use of admixtures (urea, kaolin) and different combustion systems (on a grate vs. a drop-in burner), significantly affect the energy and environmental efficiency of the combustion process. Nevertheless, in the context of the combustion process, the presence of ash is significant, as it can cause issues related to the behavior of the mineral components it contains during combustion. Chamomile ash is characterized by a much higher SiO₂ content (60.63% vs. 43.2% for ryegrass), which may explain its higher thermal stability and lower risk of slagging (FT = 1290 °C vs. 1180 °C). On the other hand, ryegrass is characterized by a higher K₂O and P₂O₅ content, which, although beneficial from the fertilizer point of view, reduces its usability as a fuel in its pure form without appropriate additives [15]. The chemical composition of ash is similar to that of herbaceous biomass, which was characterized in the work of Dyjakon [54] and Kowalczyk-Juśko [17]. The increase in the flow temperature above 1450 °C allows for safe use even in stoker boilers, exceeding the threshold of 1250 °C, which is considered safe [15]. The simultaneous use of urea and clay within these parameters resulted in a synergistic effect, which confirms the validity of modifying the composition of biofuels before combustion.

The combustion processes conducted periodically with fuel placed on the grate (type A tests) and those with automatic fuel feeding into the combustion chamber via a drop-in burner (type B tests) differ fundamentally. To ensure the most reliable testing conditions, only phase II of combustion—excluding the ignition and post-combustion stages—was analyzed. The measured combustion parameters, including combustion time, CO₂ concentration in the exhaust gases, and exhaust gas temperature, fell within the relatively wide range reported in the literature for other heating devices. It should be emphasized that during the experiments, all parameters were recorded under conditions of full heat demand from the boiler [50,51,54].

The use of admixtures, especially porcelain clay (kaolin), significantly improved the combustion parameters of both types of biomass. The change in the fuel supply system was equally important. According to the data in the work of Dula et al. [16], the use of a drop-in burner led to a significant improvement in the stability of the combustion process, which translated into an increase in efficiency (CEI increased to 87% for ryegrass) and a decrease in exhaust toxicity (TI dropped to approx. 2% at 10% O₂) [16]. CO emission decreased several times from 8–12 g·m⁻³ in system A to ~2 g·m⁻³ in system B with the addition of kaolin. This is caused by limiting the amount of fuel being burned at any given time in the burner, which helps reduce the occurrence of accidental combustion and ignition zones on the grate. These zones result from incomplete distribution of high temperatures and the movement of air through the fuel bed.

This, in turn, results in the formation of zones of incomplete combustion on the grate, which leads to the production of CO [55]. It should be noted, however, that automation of the process may lead to an increase in NO emissions, which indicates the need for further optimization of combustion parameters, especially for herbaceous biomass rich in nitrogen [16]. In the context of the data contained in work assessing the possibilities of pelleting herbaceous biomass [39], it is worth noting that the addition of kaolin reduces the density and mechanical strength of pellets and, at the same time, reduces the energy consumption in the process of their production by as much as 50%. In the context of small heating installations in rural areas, such a compromise when choosing a fuel may have significant utility and economic significance.

In comparison with the results of research [16] on the course and emissions from the combustion of other types of herbaceous and woody biomass using the same combustion methods, the following has been demonstrated.

• The combustion process of pellets made of chamomile and English ryegrass biomass together with the additives of kaolin and urea solution, during type B tests, which were described based on the CO₂ content in the exhaust gases and the temperature of the exhaust gases, did not differ significantly from similarly conducted tests using pellets made of wheat, rye, oat straw, meadow hay, and birch sawdust biomass.
• During type B tests, CO emission during the combustion of pellets with the addition of kaolin was comparable to the emission of this compound during the combustion of pellets made of biomass from wheat, rye, oat straw, meadow hay, and birch sawdust, and it was slightly lower, while for pellets without the addition of kaolin these values were even more than twice as high.
• During type A and B tests, NO emissions during the combustion of pellets with added urea solution or kaolin were lower than during the combustion of pellets made from wheat straw, rye straw, oat straw, and meadow hay biomass. However, pellets with added urea solution emitted slightly more NO than birch sawdust pellets during type B tests.
• As a rule, during type A and B tests, SO₂ emission during the combustion of pellets made of rye grass biomass with the addition of urea solution or kaolin was comparable to emissions during the combustion of pellets made of wheat straw, rye straw, meadow hay, and wood sawdust; only during the combustion of pellets in mixtures with chamomile biomass were these values higher and comparable to the emission of this compound for oat straw.
• Referring to the results of the CEI and TI parameters, the values obtained were generally comparable to the corresponding values during the combustion of pellets from biomass of wheat, rye, oat straw, meadow hay, and birch sawdust. However, during the combustion of pellets with the addition of kaolin in type B tests, the TI values were the lowest.

In summary, the comparison of the obtained results with literature data, including previous preliminary in-house studies [16,39], showed that chamomile biomass has better chemical and thermal properties. However, with appropriate modifications, ryegrass can also serve as a valuable fuel. The method of combustion and the use of additives that stabilize ash composition and reduce harmful gas emissions are of key importance. It is also necessary to expand research on the ignition process of such biomass. In this regard, it should be noted that under high-temperature conditions (>300 °C), urea decomposes, leading to the formation of ammonia (NH₃) and isocyanate (HNCO), followed by nitrogen oxides, cyanates, and other reactive nitrogen compounds. Some of these products, especially in the presence of steam and chlorine, can contribute to the creation of a highly corrosive environment (e.g., by forming HCl and NH₄Cl), which poses a threat to steel components of boiler installations. The literature [56,57,58] indicates accelerated peroxide and chloride corrosion in the presence of these compounds, particularly in the temperature range of 300–600 °C. Additionally, some of these compounds in the form of ammonia may be emitted into the atmosphere, constituting a disadvantage of such solutions [59,60].

5. Conclusions

This study of biofuels in the form of pellets from herbaceous biomass—chamomile and English ryegrass—has shown that their chemical properties are comparable to other types of herbaceous biomass that dominate the market, such as cereal straw. However, their suitability in low-power heating devices in relation to the emission requirements of such devices depends on a number of fuel modifications or process parameters. Nevertheless, based on the conducted studies, the following conclusions can be formulated:

• Pellets with kaolin (porcelain clay) do not meet the requirements for mechanical durability (PN-EN ISO 17225-6 [53]) but have better combustion parameters.
• The chemical composition of chamomile ash (higher SiO₂ content) indicates greater resistance to slagging than in the case of ryegrass (higher K₂O and P₂O₅, which is beneficial for fertilization but unfavorable during combustion).
• Type B tests showed better energy efficiency (CEI up to 87%) and lower exhaust toxicity (TI ~2%).
• The addition of kaolin reduced CO emissions by more than four times during the B-type tests, thanks to a more stable combustion process.
• However, automation of the combustion process may lead to an increase in NO emissions, which requires further optimization, especially with nitrogen-rich biomass.
• The combustion of chamomile and ryegrass pellets with additives (kaolin, urea) did not differ significantly from the results obtained for straw (wheat, rye, oat), hay, and birch sawdust in terms of CO₂ and exhaust gas temperature.
• CO and NO emissions were lower with additives, while SO₂ emissions were comparable—the exception was mixtures with chamomile (higher SO₂).

Hence, a practical conclusion can be formulated:

• Herbaceous biomass pellets, especially after modification with mineral additives and the use of an appropriate combustion system, can be an effective and relatively clean source of energy, especially in the context of small heating installations (e.g., rural ones), and such a compromise can be economically and technologically beneficial. However, further optimization of the process is necessary, especially in terms of NO emissions and pellet durability.
• In critically approaching the issue, it is also important to verify the presence of ammonia in the exhaust gases and the impact of urea on the durability of the heating system’s components.